Dear PECVD Nitride Bashers and Friends,
I asked Chris Storment, our process development expert, to compile a short note
on our current PECVD nitride process and thus answer (hopefully) questions my
response to the BBS posting started...
STS depositions conditions are:
300C
650mTorr
2000sccm SiH4(2%)in N2
35 sccm NH3
RF conditions for low stress nitrides:
High Freq. on time 6 sec (13.57MHz)
Low Freq. on time 2 sec (187.5KHz)
These RF switching times will give films that have slight tensile stresses
around 50-75MPa. Refractive Index = 2.0 +/-.05, and dep rate of about 100A/min
These films exibit BOE etch rates of about 200A/min KOH etch rates of 50A/8hrs
have been measured for KOH solutions of 20W% at 80C Typically pinholes occur
near edges of wafers with the backside (roughened side) having more pinholes.
Counts of less than 10 pinholes on the polished side are typical, while the
backside has more like 20 most near the edges of the wafer.
Pinhole problems with PECVD nitride over Al exist! When Al is heated it tends
to for hillocks at temperatures above 250C. Our nitrides go down at 300C and
hillocks have been found to form within the first 5 min of heating. After the
deposition the wafers are cooled and it is assumed that the Al punches out a
hole in the nitride and gives rise to a pinhole which exposes the Al to
etchants. We feel that this is the mechanism since other metals that are
exposed on the same substrate like Au Cr Ir Ti/W do not exhibit this problem
with pinhole formation.
Additionally we have found that pinhole counts (formed after 8hrs in 80C KOH)
can be considerable reduced when chamber conditions are carefully controlled by
using cleaning conditions that minimize particulate formation on all surfaces
of the chamber. Additionally, wafers precoated with a thermal oxide and then
coated with our PECVD nitride have lower pinhole counts after 8hr KOH etch
tests.
We hope that helps!
Best Regards,
Greg Kovacs & Chris Storment
Stanford University